Diabetes and Air Pollution

The prevalence of diabetes has risen substantially in the past decade. This increase has been linked to an “epidemic” of obesity (1). However, environmental toxins, most notably dioxins, have also been suggested as contributing factors.

Because direct exposure data, in the form of measured levels of toxicants in blood, etc., are not yet available from large populations, total air toxicants from the most recent Toxics Release Inventory (TRI) were used to evaluate the relationship between the prevalence of diabetes and environmental toxicants.

Data from the 2000 Behavioral Risk Factor Surveillance System (BRFSS) were used to determine diabetes prevalence, by state (1). These data are based on self-reports obtained from 184,450 randomly dialed participants. State total air releases for all industries from the 1999 TRI were downloaded from the U.S. Environmental Protection Agency Web site (www.epa.gov). The total prevalence of diabetes in the 2000 BRFSS was 7.3 ± 0.12% (mean ± SE)—a 49% increase from 1990 (1). Alaska had the lowest rate, 4.4%, while Mississippi had the highest, 8.8%. Race and educational level were important risk factors, with 11.1% of blacks and 12.9% of those with less than a high school education reporting diabetes.

Total reported TRI air emissions in 1999 were 2,036,510,557 lbs. Ohio industries had the highest emissions, 147,395,113 lbs, while industries in Vermont released 153,161 lbs of toxicants. Approximately 650 chemicals released by a wide variety of industries are included in the TRI. Since the reporting thresholds are high (generally 10,000 lbs) and not all industries are covered, substantial amounts of toxic chemicals are released in addition to those included in the TRI data. Dioxins and persistent bioaccumulating toxins will be included in future TRI data. (TRI data for 2001 were released in May 2000.) The dioxin reporting threshold is 0.1 g (64 FR 58666). Reporting thresholds for other persistent bioaccumulating toxins, such as pesticides, are set between 10 and100 lbs, depending on the chemical (64 FR 58666).

A linear regression analysis (Systat, Evanston, IL) revealed a significant relationship between TRI air releases, by states, and the prevalence of diabetes (r = 0.54, P = 0.000057). Although there is a large gap between air emissions and exposure, and even though the correlation between air emissions and the prevalence of diabetes does not prove a cause-and-effect relationship, the significance of the relationship demands attention.

Several studies have suggested that exposure to dioxins may be related to the development of diabetes or altered insulin metabolism (2,3,4). Dioxins are formed during the combustion of plastics, particularly in municipal and medical waste incinerators (5). Dioxins are concentrated in body fat; thus, obese individuals are likely to have an increased dioxin body burden. While dioxins are among the most toxic of all known chemicals, they are not yet included in TRI data or the Center for Disease Control’s National Report on Human Exposure to Environmental Chemicals.

I hope that this demonstration of a highly significant correlation between the prevalence of diabetes and the release of toxicants into the air will stimulate additional research in this area and lead to improvements in health.